Multi-dimensional fire detection method for locomotive

Abstract

A multi-dimensional fire detection method for a locomotive, comprising: obtaining fire characteristic data in a reference application scenario and preprocessing said fire characteristic data, and obtaining average values of the fire characteristic data in initial working states in the reference application scenario and an actual application scenario; obtaining effective first fire characteristic data in the actual application scenario and a weight coefficient of the first fire characteristic data in the actual application scenario; obtaining effective second fire characteristic data; and finally, obtaining a system fire alarm output determination parameter W to provide a basis for locomotive staff to determine whether a fire alarm has occurred.

Description

A multi-dimensional fire detection method for locomotives Technical Field

[0001] The present invention relates to the field of rail transit equipment safety technology, and in particular to a multi-dimensional fire detection method for locomotives. Background Art

[0002] Currently, relatively mature fire detection technologies and related products are available both domestically and internationally, such as temperature, smoke, and flame detectors. These devices utilize flame temperature, smoke concentration, and optical properties to detect fires, respectively. Fire detection is then determined using simple threshold algorithms based on these single detected parameters. However, in practice, fire prevention and control systems are prone to high false alarm rates, long detection times, and failures. These issues hinder effective prevention and control of vehicle fires, posing a risk to vehicle safety. Furthermore, traditional temperature, smoke, and flame detection technologies have significant drawbacks. These devices are affected by the complex application scenarios of rail transit, such as installation location, room height, humidity, dust, air velocity, oil and gas, and light, making effective fire detection difficult or impossible.

[0003] Summary of the Invention

[0004] The present invention provides a locomotive multi-dimensional fire detection method to overcome the above technical problems.

[0005] In order to achieve the above object, the technical solution of the present invention is:

[0006] A multi-dimensional fire detection method for a locomotive comprises the following steps:

[0007] S1: In a baseline application scenario, the system detection unit collects fire characteristic data of the detection area in real time; the fire characteristic data includes first fire characteristic data and second fire characteristic data, the first fire characteristic data including CO gas concentration Ct, temperature Tt, and smoke concentration; the second fire characteristic data including light signals;

[0008] S2: The system control unit preprocesses the first fire characteristic data to obtain an average value of the first fire characteristic data in the initial working state of the benchmark application scenario according to a method for obtaining an average value of the fire characteristic data in the initial working state;

[0009] S3: Obtaining the average value of the fire characteristic data in the initial working state of the actual application scenario according to the method for obtaining the average value of the fire characteristic data in the initial working state, so as to obtain valid first fire characteristic data for the actual application scenario; and simultaneously obtaining a weight coefficient of the first fire characteristic data for the actual application scenario according to the average value Aver of the fire characteristic data in the initial working state of the benchmark application scenario;

[0010] S4: The system control unit preprocesses the second fire characteristic data to obtain valid second fire characteristic data;

[0011] S5: performing normalization processing on the valid first fire characteristic data and the valid second fire characteristic data of the actual application scenario to obtain the normalized valid first fire characteristic data and the normalized valid second fire characteristic data;

[0012] S6: Based on the weight coefficient of the first fire characteristic data of the actual application scenario, the normalized effective first fire characteristic data and the normalized effective second fire characteristic data, the system fire alarm output judgment parameter W is obtained to provide a basis for the locomotive staff to judge whether a fire has occurred.

[0013] Furthermore, in S3, the method for obtaining the weight coefficient of the first fire characteristic data of the actual application scenario is as follows:

[0014] S31: Obtaining the change ratio of the average value of the first fire characteristic data: Esm=Aver_am / Aver_a1 Etm=Aver_tm / Aver_t1 Ecm=Aver_cm / Aver_c1

[0015] Wherein: Esm is the change rate of the average value of smoke concentration; Etm is the change rate of the average value of temperature Tt; Ecm is the change rate of the average value of CO gas concentration; Aver_am is the average value of smoke concentration in the initial working state of the actual application scenario; Aver_tm is the average value of temperature Tt in the initial working state of the actual application scenario; Aver_cm is the average value of CO gas concentration in the initial working state of the actual application scenario; Aver_a1 is the average value of smoke concentration in the initial working state of the benchmark application scenario; Aver_t1 is the average value of temperature Tt in the initial working state of the benchmark application scenario; Aver_c1 is the average value of CO gas concentration in the initial working state of the benchmark application scenario;

[0016] S32: Obtain the weight coefficient of the first fire characteristic data in the actual application scenario according to the change ratio of the average value of the first fire characteristic data: Am=A*Esm Tm=T*Etm Cm=C*Ecm

[0017] Where: Am is the weight coefficient of the smoke concentration in the actual application scenario; Tm is the weight coefficient of the temperature Tt in the actual application scenario; Cm is the weight coefficient of the CO gas concentration in the actual application scenario; A is the weight coefficient of the smoke concentration in the benchmark application scenario; T is the weight coefficient of the temperature Tt in the benchmark application scenario; C is the weight coefficient of the CO gas concentration in the benchmark application scenario.

[0018] Furthermore, in said S2, the steps of the method for obtaining the average value of the first fire characteristic data are as follows:

[0019] S21: within the system startup time threshold t after the system is powered on, obtain the first fire characteristic data of the system at the current moment and the first fire characteristic data of the previous moment; obtain the absolute value of the difference between the first fire characteristic data of the system at the current moment and the first fire characteristic data of the previous moment: wDiff = |nDet_Data-nTemp_Buffer|;

[0020] Wherein, nDet_Data is the first fire characteristic data received by the system at the current moment, and nTemp_Buffer is the first fire characteristic data recorded at the previous moment;

[0021] S22: When wDiff is less than the set maximum difference value DIFF_VALUE_MAX, the first fire characteristic data of the system at the current moment is recorded;

[0022] S23: According to the recorded first fire characteristic data at the current moment, an average value Aver of the first fire characteristic data in the initial working state within the system startup time threshold t is obtained.

[0023] Furthermore, in S3, the method for obtaining effective first fire characteristic data in an actual application scenario is as follows:

[0024] After the system startup time threshold t after the system is turned on, when the average value of the fire characteristic data of the initial working state of the actual application scenario is not greater than the set detection threshold mThreshold of the system detection unit, the first fire characteristic data collected by the system detection unit after the system startup time threshold t is the valid first fire characteristic data.

[0025] Furthermore, in S4, the method for preprocessing the second fire characteristic data is as follows:

[0026] S41: If the flame detection device does not obtain a light signal within the second startup time threshold after the system is turned on, execute S42; otherwise, determine that the flame detection device is faulty;

[0027] S42: After the second startup time threshold after the system is turned on, the flame detection device obtains the light signal and issues the first fire alarm, and the system control unit controls the flame detection device to power off and reset;

[0028] S43: If the flame detection device obtains a light signal for a second time after power failure and reset, and the duration of the light signal obtained for the second time is greater than the fire alarm signal duration threshold, then the light signal Ft obtained for the second time is valid second fire characteristic data.

[0029] Furthermore, in S6, the system fire alarm output judgment parameter W is obtained as follows: W = Am*a1+Tm*tem1+Cm*c1+F*f1

[0030] Where: F is the weight coefficient of the light signal Ft of the benchmark application scenario; a1 is the normalized effective smoke concentration data; tem1 is the normalized effective temperature data; c1 is the normalized effective CO gas concentration data; f1 is the normalized effective light signal data; Am is the weight coefficient of the smoke concentration in the actual application scenario; Tm is the weight coefficient of the temperature Tt in the actual application scenario; Cm is the weight coefficient of the CO gas concentration in the actual application scenario.

[0031] Beneficial Effects: The multi-dimensional fire detection method for locomotives of the present invention obtains fire characteristic data under a benchmark application scenario, pre-processes it, and obtains the average value of the fire characteristic data of the initial working state of the benchmark application scenario and the actual application scenario; obtains the effective first fire characteristic data of the actual application scenario and the weight coefficient of the first fire characteristic data of the actual application scenario; simultaneously obtains the effective second fire characteristic data; and finally obtains the system fire alarm output judgment parameter W to provide a basis for locomotive staff to judge whether a fire has occurred. According to the different application scenarios of rail transit and the reasonable arrangement of fire detectors in these scenarios, the multi-source information obtained by various sensors can be comprehensively integrated, effectively solving the problem of high false alarm rate and detection failure of traditional detectors due to interference from the external environment, improving the timeliness and accuracy of fire identification, and providing a strong guarantee for the safe and stable operation of vehicles. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0033] FIG1 is a flow chart of a fire detection method according to the present invention;

[0034] FIG2 is a schematic diagram of an application scenario of a fire detection system and detector selection in an embodiment of the present invention;

[0035] FIG3 is a schematic diagram of a fire characteristic data preprocessing process in an embodiment of the present invention;

[0036] FIG4 is a flow chart of a fire detection method in an embodiment of the present invention. DETAILED DESCRIPTION

[0037] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0038] This embodiment provides a multi-dimensional fire detection method for a locomotive, as shown in FIG1 , including the following steps:

[0039] In the embodiments of the present invention, detectors are first selected based on the application scenario of the fire detection system. Specifically, the application scenarios include detection areas such as the locomotive electrical room, diesel engine room, electrical cabinet, and battery cabinet. Different types of detectors are selected according to different detection areas. The detectors mainly include smoke detectors, temperature-sensitive flame detectors, characteristic gas detectors, and aspirating detectors. Photoelectric smoke detectors and temperature detectors are used for detection in the electrical room; photoelectric smoke detectors, temperature detectors, and flame detectors are used for detection in the diesel engine room; aspirating detectors (combined detection of smoke and CO gas) are used for detection in the electrical cabinet; and temperature detectors and CO characteristic gas detectors are used for detection in the battery cabinet. This is shown in Figure 2.

[0040] S1: In a benchmark application scenario, the system detection unit collects fire characteristic data of the detection area in real time; the fire characteristic data includes first fire characteristic data and second fire characteristic data, the first fire characteristic data includes CO gas concentration Ct, temperature Tt, and smoke concentration; the second fire characteristic data includes light signal Ft; specifically, the benchmark application scenario can be a test scenario of the system detection unit before leaving the factory, or it can be other scenarios, which serves as a standard for calculating the weight coefficients of CO gas concentration data, smoke concentration data and temperature when the environment changes.

[0041] Specifically, as shown in Figure 3, the detector transmits detected fire characteristic parameters such as smoke, temperature, CO characteristic gas concentration, and flame light signals to the controller, which performs data preprocessing on these different fire characteristic parameters. Communication between the controller and the detector occurs via a dual bus, RS485, and CAN communication. Dynamic data storage and filtering are performed on smoke, temperature, and characteristic gas concentration parameters. After operating for a period of t, the system calculates the average value of each dynamically stored characteristic parameter and uses this average value to determine if the detector is operating properly. For light signal characteristic parameters, the system performs data preprocessing based on the detector's fire alarm output time and frequency to prevent false alarms from the flame detector.

[0042] S2: The system control unit preprocesses the first fire characteristic data to obtain an average value Aver of the first fire characteristic data in the initial working state of the benchmark application scenario according to a method for obtaining an average value of the fire characteristic data in the initial working state;

[0043] Preferably, the steps of the method for obtaining the average value of the first fire characteristic data in the initial working state are as follows: S21: within a system startup time threshold t after the system is powered on, obtaining the first fire characteristic data of the system at the current moment and the first fire characteristic data of the previous moment; and obtaining the absolute value of the difference between the first fire characteristic data of the system at the current moment and the first fire characteristic data of the previous moment: wDiff = |nDet_Data - nTemp_Buffer|;

[0044] Wherein, nDet_Data is the first fire characteristic data received by the system at the current moment, and nTemp_Buffer is the first fire characteristic data recorded at the previous moment;

[0045] S22: When wDiff is less than the set maximum difference value DIFF_VALUE_MAX, the first fire characteristic data of the system at the current moment is recorded;

[0046] S23: Based on the first fire characteristic data recorded at the current moment, obtain the average value Aver of the first fire characteristic data of the initial working state within the system startup time threshold t; wherein, the method for obtaining the average value Aver of the first fire characteristic data of the initial working state within the system startup time threshold t is a conventional method for calculating the average value in the field and is not described in detail here.

[0047] S3: Obtaining the average value of the fire characteristic data in the initial working state of the actual application scenario according to the method for obtaining the average value of the fire characteristic data in the initial working state, so as to obtain valid first fire characteristic data for the actual application scenario; and simultaneously obtaining a weight coefficient of the first fire characteristic data for the actual application scenario according to the average value Aver of the fire characteristic data in the initial working state of the benchmark application scenario;

[0048] Preferably, the method for obtaining the weight coefficient of the first fire characteristic data in an actual application scenario is as follows:

[0049] S31: Obtain the change ratio of the average value of the first fire characteristic data: Esm = Aver_am / Aver_a1 Etm = Aver_tm / Aver_t1 Ecm = Aver_cm / Aver_c1

[0050] Wherein: Esm is the change rate of the average value of smoke concentration; Etm is the change rate of the average value of temperature Tt; Ecm is the change rate of the average value of CO gas concentration; Aver_am is the average value of smoke concentration in the initial working state of the actual application scenario; Aver_tm is the average value of temperature Tt in the initial working state of the actual application scenario; Aver_cm is the average value of CO gas concentration in the initial working state of the actual application scenario; Aver_a1 is the average value of smoke concentration in the initial working state of the benchmark application scenario; Aver_t1 is the average value of temperature Tt in the initial working state of the benchmark application scenario; Aver_c1 is the average value of CO gas concentration in the initial working state of the benchmark application scenario;

[0051] S32: Obtain the weight coefficient of the first fire characteristic data in the actual application scenario according to the change ratio of the average value of the first fire characteristic data: Am=A*Esm Tm=T*Etm Cm=C*Ecm

[0052] Where: Am is the weight coefficient of the smoke concentration in the actual application scenario; Tm is the weight coefficient of the temperature Tt in the actual application scenario; Cm is the weight coefficient of the CO gas concentration in the actual application scenario; A is the weight coefficient of the smoke concentration in the benchmark application scenario; T is the weight coefficient of the temperature Tt in the benchmark application scenario; C is the weight coefficient of the CO gas concentration in the benchmark application scenario.

[0053] Preferably, the method for obtaining effective first fire characteristic data in actual application scenarios is as follows:

[0054] After the system startup time threshold t after the system is turned on, when the average value of the fire characteristic data of the initial working state of the actual application scenario is not greater than the set detection threshold mThreshold of the system detection unit, the first fire characteristic data collected by the system detection unit after the system startup time threshold t is valid first fire characteristic data; otherwise, it is determined that the sensor corresponding to the first fire characteristic data of the system detection unit is faulty, and the detection value of the faulty sensor is no longer used as the basis for fire alarm judgment.

[0055] In this embodiment, when the absolute value wDiff of the difference between the first fire characteristic data at the current moment and the first fire characteristic data at the previous moment is greater than the set maximum difference DIFF_VALUE_MAX, the first fire characteristic data at the current moment is an interference signal.

[0056] Specifically, after the system is turned on and within the system startup time threshold t, the fire alarm algorithm judgment is not performed on the detection data of the first fire characteristic data, and only the first fire characteristic data is preprocessed. First, the control unit opens up a data cache space to store various characteristic number parameters. Suppose the space size of a certain characteristic data parameter is N, and m data can be stored in real time. The data is stored in a queue manner. Furthermore, during the real-time data reception process, the first fire characteristic data is filtered. nDet_Data is defined as the detector characteristic parameter at a specific moment, nTemp_Buffer is the characteristic parameter recorded at the previous moment, and wDiff = |nDet_Data - nTemp_Buffer| is defined as the absolute value of the difference between the two. A determination is made as to whether wDiff is greater than a set maximum difference value, DIFF_VALUE_MAX. DIFF_VALUE_MAX is set by the application scenario. If wDiff is greater than the set maximum difference value, nDet_Data is considered an interference signal at the current moment and is not stored. Otherwise, nDet_Data is considered normal data and is stored. nDet_Data is also updated to nTemp_Buffer for filtering at the next moment. The system startup time threshold, t, after system startup is manually set and varies between detectors from different manufacturers.

[0057] Furthermore, after the system is powered on and the system startup time threshold t is reached, the average value Aver of the initial working state of various first fire characteristic data is calculated. When the average value Aver of a certain first fire characteristic data is greater than the threshold mThreshold set under the normal working state (mThreshold is set by the application scenario), it is determined that the sensor collecting the first fire characteristic data has failed, and the characteristic parameters of the sensor are no longer used as the basis for fire alarm judgment.

[0058] S4: The system control unit preprocesses the second fire characteristic data to obtain valid second fire characteristic data;

[0059] Preferably, the method for preprocessing the second fire characteristic data is as follows:

[0060] Specifically, the system's detection unit collects characteristic parameters of the light signal Ft in real time. The light signal Ft received by the system is a dry contact signal, that is, a switch signal of 0 or 1. If the detector detects a flame, it outputs a signal of 1, which is a fire alarm signal; if no flame is detected, it outputs a signal of 0. To prevent the flame detector from falsely reporting a fire alarm, the flame detector data processing method is as follows:

[0061] S41: If the flame detection device does not obtain a light signal within the second startup time threshold after the system is turned on, execute S42; otherwise, the flame detection device fails;

[0062] S42: After the second startup time threshold after the system is turned on, the flame detection device obtains the light signal and issues the first fire alarm, and the system control unit controls the flame detection device to power off and reset;

[0063] S43: When the flame detection device obtains a light signal for a second time after power failure and reset, and the duration of the light signal obtained for the second time is greater than the fire alarm signal duration threshold, the light signal Ft obtained for the second time is valid second fire characteristic data.

[0064] Specifically, in this embodiment, when the system is first powered on and within the second startup time threshold of 10 seconds, if the flame detector is determined to have issued a fire alarm, the system deems the flame detector faulty and no longer uses the detector's optical signal characteristic parameters as a basis for fire alarm determination. 10 seconds later, the flame detector begins normal operation. If the flame detector obtains an optical signal and issues a first fire alarm, the control unit powers off and resets the flame detector. If the detector obtains an optical signal a second time after the reset and issues a fire alarm again, and the duration of the second optical signal exceeds the fire alarm signal duration threshold, i.e., the flame detector fire alarm signal persists for more than 5 seconds, the system control unit determines that the flame detector has issued a fire alarm. The optical signal Ft output by the flame detector is valid second fire characteristic data.

[0065] S5: performing normalization processing on the effective first fire characteristic data and the effective second fire characteristic data of the actual application scenario to obtain the normalized effective first fire characteristic data and the normalized effective second fire characteristic data;

[0066] Specifically, the pre-processed Ct, Tt, St, and Ft fire characteristic data are normalized and mapped to a range from 0 to 1. The normalization method is a conventional technique in the field and will not be described in detail here.

[0067] S6: Based on the weight coefficient of the first fire characteristic data of the actual application scenario, the normalized effective first fire characteristic data and the normalized effective second fire characteristic data, the system fire alarm output judgment parameter W is obtained to provide a basis for the locomotive staff to determine whether a fire has occurred.

[0068] Preferably, the system fire alarm output judgment parameter W is obtained as follows: W = Am*a1+Tm*tem1+Cm*c1+F*f1

[0069] Where: F is the weight coefficient of the light signal Ft of the benchmark application scenario; a1 is the normalized effective smoke concentration data; tem1 is the normalized effective temperature data; c1 is the normalized effective CO gas concentration data; f1 is the normalized effective light signal data;

[0070] Specifically, the system communicates with the locomotive central control unit through a network, and the communication method adopts Ethernet or CAN communication. The system can obtain the locomotive operating condition, time, altitude, geographical location and other status information in real time. By obtaining the status information of the locomotive under different operating environments, the system can dynamically adjust the characteristic weight coefficients Am, Tm, and Cm. The adjustment method is: when a certain environmental factor changes, such as the locomotive operating condition, the average value Aver of the fire characteristic data of the initial working state of the benchmark application scenario is used to recalculate the average value of the fire characteristic data of the initial working state of the various characteristic parameters in the actual application scenario, which are the average value Aver_am of the smoke concentration in the initial working state of the actual application scenario, the average value Aver_tm of the temperature Tt in the initial working state of the actual application scenario, and the average value Aver_cm of the CO gas concentration in the initial working state of the actual application scenario;

[0071] Specifically, in traditional fire alarm judgment methods, the system uses a weighted algorithm to determine whether a fire has occurred. Depending on the application scenario, different weight coefficients are assigned to multiple sensor parameters. These coefficients primarily include the weight coefficient A for smoke concentration in the benchmark application scenario, the weight coefficient T for temperature Tt in the benchmark application scenario, the weight coefficient C for CO gas concentration in the benchmark application scenario, and the weight coefficient F for light signal Ft in the benchmark application scenario. The system fire alarm output judgment parameter is W. W = A*s1 + T*tem1 + C*c1 + F*f1, where the weight coefficients for each parameter in the benchmark application scenario are determined based on the benchmark application scenario and surrounding environmental factors, based on the staff's experience. Finally, the W value is used to determine whether the system has issued an alarm. The system fire alarm threshold is set to Wt. If W>Wt, the system is deemed to have issued a fire alarm; otherwise, no fire alarm has occurred, as shown in Figure 4.

[0072] In an embodiment of the present invention, the average values ​​of smoke concentration, temperature Tt, and CO gas concentration in the initial working state of the benchmark application scenario before the environmental factors change are Aver_a1, Aver_t1, and Aver_c1 respectively; the change ratio of the average values ​​of each first fire characteristic data is calculated, and the weight coefficient of each characteristic data is adjusted due to different application environments. The adjusted characteristic weight coefficient is Am=A*Esm, Tm=T*Etm, Cm=C*Ecm; the system fire alarm output is weighted as W=Am*a1+Tm*tem1+Cm*c1+F*f1. The dynamically adjusted weight coefficient based on the benchmark application scenario can make the fire alarm output judgment parameter W value more realistic and guiding.

[0073] Specifically, this embodiment selects detectors based on actual application scenarios, and the system performs data preprocessing and data fusion on the detected fire characteristic data. The system can obtain the locomotive operating conditions in real time, and the controller can dynamically adjust the algorithm feature weight coefficient based on the status information of the locomotive under different operating environments. The system collects fire characteristic data in real time and calculates the fire alarm output judgment parameter W value based on the feature weight coefficient. When the system determines that a fire has occurred, the fire controller emits an audible and visual alarm and uploads the fire alarm signal to the TCMS system (locomotive monitoring and control system); it controls the solenoid valve on the cylinder to open, release the fire extinguishing agent, and suppress the fire.

[0074] Beneficial Effects: The multi-dimensional fire detection method for locomotives of the present invention obtains fire characteristic data under a benchmark application scenario, pre-processes it, and obtains the average value of the fire characteristic data of the initial working state of the benchmark application scenario and the actual application scenario; obtains the effective first fire characteristic data of the actual application scenario and the weight coefficient of the first fire characteristic data of the actual application scenario; simultaneously obtains the effective second fire characteristic data; and finally obtains the system fire alarm output judgment parameter W to provide a basis for locomotive staff to judge whether a fire has occurred. According to the different application scenarios of rail transit and the reasonable arrangement of fire detectors in these scenarios, the multi-source information obtained by various sensors can be comprehensively integrated, effectively solving the problem of high false alarm rate and detection failure of traditional detectors due to interference from the external environment, improving the timeliness and accuracy of fire identification, and providing a strong guarantee for the safe and stable operation of vehicles.

[0075] This embodiment utilizes different sensor configurations based on the environment and requirements of the detection site, leveraging the strengths of each sensor and leveraging their strengths to achieve multi-information fusion detection with shared information. By filtering and preprocessing the detection data, interference and fault signals are eliminated, making the detection data more accurate and reliable. Multi-dimensional simultaneous detection is implemented to address the influence of different scenarios and environmental factors, integrating data from different types of fire characteristic parameters. Networking with the locomotive central control unit allows for acquisition of locomotive status information and dynamic adjustment of characteristic weight coefficients. This makes the fire alarm algorithm more intelligent, effectively reducing the system's false alarm rate and enabling accurate fire identification.

[0076] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A locomotive multi-dimensional fire detection method, characterized in that: The steps include: S1: In the benchmark application scenario, the system detection unit collects fire characteristic data of the detection area in real time; the fire characteristic data includes first fire characteristic data and second fire characteristic data, the first fire characteristic data includes CO gas concentration Ct, temperature Tt, and smoke concentration; the second fire characteristic data includes light signals; S2: the system control unit preprocesses the first fire characteristic data to obtain an average value of the first fire characteristic data in the initial working state of the reference application scenario according to a method for obtaining an average value of the fire characteristic data in the initial working state; S3: according to the method for obtaining the average value of the fire characteristic data in the initial working state, obtaining the average value of the fire characteristic data in the initial working state of the actual application scenario, so as to obtain the effective first fire characteristic data of the actual application scenario; at the same time, according to the average value Aver of the fire characteristic data in the initial working state of the benchmark application scenario, obtaining the weight coefficient of the first fire characteristic data of the actual application scenario; S4: the system control unit preprocesses the second fire characteristic data to obtain valid second fire characteristic data; S5: normalizing the valid first fire characteristic data and the valid second fire characteristic data of the actual application scenario to obtain the normalized valid first fire characteristic data and the normalized valid second fire characteristic data; S6: According to the weight coefficient of the first fire characteristic data of the actual application scenario, the normalized effective first fire characteristic data and the normalized effective second fire characteristic data, the system fire alarm output judgment parameter W is obtained to provide a basis for locomotive staff to judge whether a fire has occurred.

2. A locomotive multi-dimensional fire detection method according to claim 1, characterized in that: In S3, the method for obtaining the weight coefficient of the first fire characteristic data of the actual application scenario is as follows: S31: Obtaining the change ratio of the average value of the first fire characteristic data: Esm=Aver_am / Aver_a1 Etm=Aver_tm / Aver_t1 Ecm=Aver_cm / Aver_c1 Wherein: Esm is the change ratio of the average value of smoke concentration; Etm is the change ratio of the average value of temperature Tt; Ecm is the change ratio of the average value of CO gas concentration; Aver_am is the average value of smoke concentration in the initial working state of the actual application scenario; Aver_tm is the average value of temperature Tt in the initial working state of the actual application scenario; Aver_cm is the average value of CO gas concentration in the initial working state of the actual application scenario; Aver_a1 is the average value of smoke concentration in the initial working state of the benchmark application scenario; Aver_t1 is the average value of temperature Tt in the initial working state of the benchmark application scenario; Aver_c1 is the average value of CO gas concentration in the initial working state of the benchmark application scenario; S32: Obtaining a weight coefficient of the first fire characteristic data of an actual application scenario according to a change ratio of the average value of the first fire characteristic data: Am=A*Esm Tm=T*Etm Cm=C*Ecm In the formula: Am is the weight coefficient of the smoke concentration in the actual application scenario; Tm is the weight coefficient of the temperature Tt in the actual application scenario; Cm is the weight coefficient of the CO gas concentration in the actual application scenario; A is the weight coefficient of the smoke concentration in the benchmark application scenario; T is the weight coefficient of the temperature Tt in the benchmark application scenario; C is the weight coefficient of the CO gas concentration in the benchmark application scenario.

3. A locomotive multi-dimensional fire detection method according to claim 1, characterized in that: In S2, the steps of the method for obtaining the average value of the first fire characteristic data are as follows: S21: within the system startup time threshold t after the system is turned on, obtain the first fire characteristic data of the system at the current moment and the first fire characteristic data of the previous moment; to obtain the absolute value of the difference between the first fire characteristic data of the system at the current moment and the first fire characteristic data of the previous moment: wDiff=|nDet_Data-nTemp_Buffer|; Wherein, nDet_Data is the first fire characteristic data received by the system at the current moment, and nTemp_Buffer is the first fire characteristic data recorded at the previous moment; S22: When wDiff is less than the set maximum difference value DIFF_VALUE_MAX, the first fire characteristic data of the system at the current moment is recorded; S23: According to the recorded first fire characteristic data at the current moment, an average value Aver of the first fire characteristic data of the initial working state within the system startup time threshold t is obtained.

4. A locomotive multi-dimensional fire detection method according to claim 1, characterized in that: In S3, the method for obtaining the first effective fire characteristic data of the actual application scenario is as follows: After the system startup time threshold t after the system is turned on, when the average value of the fire characteristic data of the initial working state of the actual application scenario is not greater than the set detection threshold mThreshold of the system detection unit, the first fire characteristic data collected by the system detection unit after the system startup time threshold t is the valid first fire characteristic data.

5. The locomotive multi-dimensional fire detection method according to claim 1, characterized in that: In S4, the method for preprocessing the second fire characteristic data is as follows: S41: within the second startup time threshold after the system is turned on, if the flame detection device does not obtain the light signal, execute S42; otherwise, it is determined that the flame detection device is faulty; S42: After the second startup time threshold after the system is turned on, the flame detection device obtains the light signal to report the fire alarm for the first time, and the system control unit controls the flame detection device to power off and reset; S43: If the flame detection device obtains a light signal for a second time to issue a second fire alarm after power failure and the duration of the light signal obtained for the second time is greater than the fire alarm signal duration threshold, then the light signal Ft obtained for the second time is valid second fire characteristic data.

6. A locomotive multi-dimensional fire detection method according to claim 1, characterized in that: In S6, the system fire alarm output judgment parameter W is obtained as follows: W = Am*a1+Tm*tem1+Cm*c1+F*f1 Where: F is the weight coefficient of the light signal Ft of the benchmark application scenario; a1 is the normalized effective smoke concentration data; tem1 is the normalized effective temperature data; c1 is the normalized effective CO gas concentration data; f1 is the normalized effective light signal data; Am is the weight coefficient of the smoke concentration in the actual application scenario; Tm is the weight coefficient of the temperature Tt in the actual application scenario; Cm is the weight coefficient of the CO gas concentration in the actual application scenario.

Images